PLASMID

25,64-69

( 199 1)

Identification of a Region That Influences Host Range of the Streptococcal Conjugative Plasmid plP501 E. REGIS KRAH III AND FRANCIS L. MACRINA’ Department ofklicrobiology and Immunology, Virginia Commonwealth University, Richmond, Virginia 23298-0678 Received February 2, 1990; revised November

1, 1990

pIPSO is a member of a group of conjugative plasmids that are self-transmissible to a wide variety of streptococci as well as to other gram-positive bacteria. Several pIPSO restriction fragment deletion derivatives have been isolated and characterized. In this paper we describe one such derivative (pVA 1702) which was conjugally proficient but had a limited host range. The loss of host range ability was seen as decreased conjugal transfer from Enterococcusfaecalis to Streptococcus sanguis and was coincident with the deletion of a 4.5kb DNA fragment. Transformation of pVA 1702 into S. sanguis also was dramatically reduced as compared to its progenitor, suggesting the 4.5kb fragment encoded a factor(s) necessary for stable maintenance in this host but not in E. fhecalis. These observations suggest that pIP501 employs specific 0 1991 Academic Press, Inc. mechanisms enabling its maintenance in certain gram-positive bacteria.

of streptococci (Bougueleret et al., 1981; Buu-Hoi et al., 1984; Clewell 198 1; Horaud et al., 1985; Horodniceanu et al., 1976) as well as to other gram-positive bacteria such as staphylococci (Schaberg et al., 1982), bacilli (Oultram and Young, 1985) lactobacilli (Gibson et al., 1979) pediococci (Gonzalez and Kunka, 1983), Listeriu (Buu-Hoi et al., 1984), and clostridia (Oultram and Young, 1985). Two of these plasmids, pIPSO (Behnke and Gilmore, 198 1; Evans and Macrina, 1983; Krah and Macrina, 1989) and PAM/~ 1 (LeBlanc and Lee, 1984) have been characterized at the molecular level and have been used to develop several useful gene cloning systems (Behnke and Gilmore, 198 1; Evans et al., 1985). Although these replicons and their derivatives have been used to mobilize recombinant DNA molecules to a variety of gram-positive bacteria (Romero et al., 1987), little is known of the mechanism(s) used by these plasmids to maintain themselves in this wide variety of hosts. pIP50 1 is 30.2 kb in size (Fig. 1) and carries a chloramphenicol resistance determinant in addition to the common MLS’ gene (Evans and Macrina, 198 3). The genesis of a number

Broad host range conjugative plasmids have been found in a number of gram-positive and gram-negative bacteria. Such plasmids are believed to mediate the rapid horizontal exchange of genetic information through bacterial populations. RK2, a gramnegative incP conjugative plasmid, can transfer itself and be stably maintained in more than nine different species (Schmidhauser and Helsinki, 1985). This promiscuity may be mediated by a complex system which controls plasmid replication in different host backgrounds (Young et al., 1987). In streptococci, a group of broad host range conjugal plasmids have been characterized (Clewell, 198 1). These plasmids have several common features: a conserved macrolide, lincosamide, and streptogramin B resistance (MLS’) determinant; considerable DNA sequence similarity as determined by hybridization studies; and a common size ranging from 26 to 33 kb (Horaud et al., 1985). In filter matings these plasmids transfer to a wide variety ’ To whom correspondence should be addressed to at: Box 678, MCV Station, Richmond, VA 23298-0678. FAX: (804) 786-9946. 0147-619X/91

$3.00

Copyright 0 I99 I by Academic Press, Inc. All rights of reproduction in any form reserved.

64

SHORT COMMUNICATIONS

FIG. 1.Composite map of PIP50 1.ori, origin of replication; cop,copy number control; Em, erythromycin resistance gene; and Cm, chioramphenicol resistance gene (Behnke and Gilmore, 1981, Evans and Macrina, 1983). Arrows indicate the presumed direction of transcription of the illustrated genes (Krah and Macrina, 1989). Transposon insertions into the A and B regions abolish conjugative ability (Krah and Macrina, 1989). The genes for ~67, ~79, and the conjugation (c~J] functions ofthe A region were previously described (Krah and Macrina, 1989). The light stippled region (labeled Stability) corresponds to the 4.5kb fragment that affects establishment of pIPSO derivatives in S. sang& (see text).

of PIP50 1derivatives constructed in our laboratory is shown in Fig. 2. We had prepared several deletion derivatives by in vitro removal of restriction fragments of pIP501 in order to locate the genesinvolved in conjugal transfer (tra+ genes) (Krah and Macrina, 1989). pVA1701 and pVA1702 were two conjugally proficient pIPSO1 derivatives (Krah and Macrina, 1989; Fig. 2) which promoted their self transfer in isogenic and nonisogenic Enterococcus faecalis matings. These plasmids differed by a 4.5-kb DNA fragment (Fig. 1). Surprisingly, we observed that donor strains carrying pVAl701 or pVA 1702 differed dramatically in their of transfer frequency to non-faecalis recipients. (see Table 1 and discussion below). In this paper we describe the genetic basis of these observations. Filter matings, isolation of plasmid DNA, restriction endonuclease digests, and gel electrophoresis were performed as described (Krah and Macrina, 1989). Bacteria were

65

grown in brain heart infusion broth (BHI; Difco Laboratories, Detroit, MI) with agar added to 1.5% to make solid media. Antibiotics were used in solid media at the following concentrations: streptomycin (Sm), 200 pg/ml; spectinomycin (Sp), 200 pg/ml; rifampin (Rf), 20 &ml; fusidic acid (Fs), 20 &ml; kanamycin (Km), 1000 &ml. E. faecalis strains used in this work included OG lSS(chromosomally resistant to streptomycin and spectinomycin) and OGl-RF (chromosomally resistant to rifampicin and fusidic acid). Both of these strains were provided by Don B. Clewell, University of Michigan. Streptococcussanguis Challis derivatives included V48 1 (chromosomally resistant to rifampicin) and V1828 (chromosomally resistant to rifampicin and fusidic acid). The kanamycin resistance gene on pVAl70 1 and pVA 1702 originated from plasmid pStr4A (10.4 kb) kindly provided by Donald J. LeBlanc, University of Texas Health Sciences Center, San Antonio. In isogenic E. ,faecalismatings, pVA 1702 transferred at a higher frequency than pVA1701 (Table 1; line 1). In contrast, when S. sanguis was used as the recipient, the transfer frequency of pVA1702 decreased 1 X 106-fold, whereas pVA 1701 only dropped loo-fold (Table 1, compare lines 1 and 2). The former frequency was at the lower limit of detection for conjugative transfer in our system but colonies obtained from such crosseswere uniformly found to be S. sanguis carrying a plasmid indistinguishable from pVA1702. Such progeny of the E. faecalis (donor)/S. sanguis (recipient) mating were used as donors in crosseswith an E. faecalis recipient. Both plasmids transferred at high frequency into E. faecalis (Table 2, line 3). The pVA 1702-carrying E. faecalis progeny of this mating then were used as donors in matings with S. sanguis. Surprisingly, pVA 1702 donor strains derived in this manner no longer exhibited a restricted host range and transferred at the same frequency as pVA170 1 (Table 1, lines 2 and 4). One donor strain obtained in this fashion was pre-

66

SHORT COMMUNICATIONS

P ECORI

Hpal I Hpa1 I

WA3BO-I 42kD 0

Hpall-Aval and ligation

clevage

-Y-

u

Aval

Tra+ Tn9 17 msertlon

pUA797 31 kb

Bell- EcoRl dlgestlon and replacement with Km gene of pStr-4A /

I t z

Aval dlgestion and self llgatlon Removes Tn9 I7 InsertIon and 4.5 kb of plasmld DNA

FIG. 2. Lineage of pVA 1701 and pVA 1702.Following the indicated in vitro reactions, all constructs were recovered after transformation into either E. faecalis or S. sanguis. Only relevant restriction endonuclease sitesare shown on the maps. pVA797 was constructed by the addition of the pVA380- 1 replicon to pIPSO1; this strategy resulted in the removal the Em’ marker from pIPSOl. A pVA797 transpson-carrying plasmid (Krah and Macrina, 1989), pVA1229, was used to remove an AvuI fragment, deleting all but 5 bp of the Tn917 transposon and approximately 4.5kb of plasmid DNA, thus creating pVA 1700.The Cm’gene and the pVA380- 1 origin of replication carried by pVA 1700 were removed as a 3.8-kb BclI-EcoRI fragment and the Km’ gene of pStr4A was inserted in their place, creating pVA1702. pVA 1701 was constructed directly from pVA797 using an identical restriction endonuclease cleavage strategy to replace similar sequenceswith the Km’ marker. Hence, pVA1701 and pVA1702 are effectively isogenic derivatives of pIP50 1, with pVA 1702 missing an additional 4.5 kb of contiguous DNA.

67

SHORT COMMUNICATIONS TABLE I CONJ~GATIVEKAIUMY~INRESISTANCJSTRANSFERFRFQUENCY" Donor

Recipient

Selection

E. faecalis E. faecalis S. sanguis E. faecalis

E. faecaW’ S. sanguis E. faecalis S. sanguis

Km,Rp Km, Rf, Fs Km, Sm Km, Rf

pVAl701 2.18 x lo-’ 5.11 x 10-5 1.06 x lo-’ N.D.

pVA1702

pVAl702B

1.10 x 10-z 1.77 x 10-r 1.20 x 1o-3 N.D.

N.D. N.D. N.D. 1.36 x lO-5

a Isogenic mating pair [OG 1-SS X OG I -RF (seetext)]. Similar mating frequencies were observed when mating were performed with nonisogenic E. faecalis. bAbbreviations: Km, kanamycin; Rf, rifampicin; Fs, fusidic acid, Sm, streptomycin. ’ N.D., not done.

they promoted self transfer at a frequency similar to pVA 1702B). To determine if the deletion carried by pVAl702 created a DNA fusion which was deleterious in some hosts but not others, the transformation frequencies of selected pVA 1702 : : Tn9 17 derivatives (Krah and Macrina, 1989) were determined. The transposon insertions on these plasmids flanked the &A-AvaI fragment of pVA1702 (seeFig. 2), physically separating this region from the rest of the plasmid. If the 4.5-kb deletion created a DNA fusion which destabilized the plasmid in the S. sanguis background, then an insertion into this region of the plasmid should disrupt the fusion restoring stability. The transformation frequency of such pVA 1702: : Tn9 17 derivatives into S. sanguis was similar to pVA1702 (data not shown). These data further suggestedthat the difference in host range ability was due to missing genetic information rather than the activation of plasmid genes owing to the novel fusion of plasmid sequences. The stability of pVA 1701, pVA 1702, and TABLE 2 pVA 1702B was compared by analyzing the COMPARI~ONOFTRANSFORMATIONFREQUENCY frequency of plasmid loss under nonselective 1~~0s. SANGUISV481" conditions. Samples from overnight broth pStr4A 3.5 x 10-r cultures of plasmid-carrying cells grown pVA1701 3.23 x lo-’ under selective conditions were diluted pVA 1702 1.27 x 1O-6 1:10,000 into prewarmed. drug-free medium. pVA1702B 1.61 x IO-’ Cell growth was monitored at ODhbOand, at ’ Expressed as the frequency of Km’ transformants per appropriate time points, samples were dicompetent cell at the time of DNA addition. luted and grown on nonselective solid media.

sumed to have a second-site, extrachromosomal mutation and its plasmid was designated pVA 1702B. Comparison ofmultiple restriction endonuclease digest patterns of pVA1702 and pVA1702B did not reveal detectable differences (data not shown) suggesting the latter plasmid carried a point mutation. The need for a functioning conjugation system was bypassedby comparing the transformation frequency of pVA 170 1, - 1702, and -1702B into S. sanguis. All plasmids were isolated from E. faecalis OGl-SS to control for host modification effects.One microgram of plasmid DNA in 100 ~1of TE ( 10 mM Tris, 1 mM EDTA, pH 8) was added to 500 ~1 of competent S. sanguis cells. pVA 1702 transformed V48 1 IOOO-fold less efficiently than pVA 1701 or pVA 1702B (Table 2). S. sanguis pVA1702 transformants were tested for conjugal ability and found to no longer exhibit a restricted host range (i.e.,

68

SHORT COMMUNICATIONS

For each time point, 200 colonies were picked onto kanamycin-containing medium and the number of resistant colonies divided by the total number of colonies picked was interpreted as the percentage of bacteria maintaining the plasmid. After 25 generations of growth greater than 99% of E. faecalis cells maintained pVA 1701, while approximately 95% maintained pVAl702. However, less than 25% of E. faecalis cells maintained pVA 1702B under similar conditions. Both pVA 1701 and pVA 1702B were stably maintained in S. sanguis (>99%) for more than 25 generations. Taken together our data argue that pIP50 1 has a specialized mechanism for stable plasmid maintenance in certain host backgrounds. Our experiments do not rigorously rule out the possibility the deletion of the 4.5kb fragment generated a DNA fusion which is deleterious to S. sanguis but not E. faecalis (e.g., fusion of a gene to a new promoter that is more active in one host than another). However, we feel this is highly unlikely since transformation frequencies of pVA 1702 : : Tn9 17 derivatives which flanked the BclIAvaI fragment transformed S. sanguis at frequencies similar to pVA1702. These results suggest that the 4.5-kb restriction fragment encodes some factor which is necessary for stable plasmid maintenance in some hosts but not others. This factor may function in trans on the pIP501 origin of replication to extend its host range. Alternatively, this 4.5kb region may contain a second origin of replication which is used only in some hosts. Clearly this region is not necessary for stable maintenance in E. faecalis, otherwise it would not have been possible to construct pVA 1702. Transmission of pVA 1702 into S. sanguis by either conjugation or transformation presumably selected for replicons that carried mutations which corrected the defect created by the deletion of the 4.5-kb fragment. This reacquisition of wider host range ability was not complete, for such mutant plasmids were no longer stably maintained in E. faecalis. These observations provide the

first evidence that the maintenance of plasmids like pIP501 may involve multiple genetic loci that function in some host cells and not others. ACKNOWLEDGMENTS This work was supported by USPHS Grant R37 DE04224 to F.L.M. We thank D. B. Clewell for providing us with E. faecalis strains.

REFERENCES BEHNKE,D., AND GILMORE, M. S. (198 I). Location of antibiotic resistance determinants, copy control, and replication functions on the double-selective streptococcal cloning vector pGB30 1. Mol. Gen. Genez.184, 115-120. BOUGUELERET, L., BIETH, G., AND HORODNICEANU,T. (1981). Conjugative R plasmids in group C and G streptococci. J. Bacterial. 145, 1102-I 105. Buu-HOI, A., BETH,G., AND HORAUD,T. (1984). Broad host range of streptococcal macrolide resistance plasmids. Antimicrob. Agents Chemother. 25,289-29 1. CLEWELL,D. B. (1981). Plamids, drug resistance, and gene transfer in the genus Streptococcus. Microbial. Rev. 45,409-436.

CLEWELL, D. B., YAGI, Y., DUNNY, G. M., AND SCHULTZ,S. K. (1974). Characterization of three plasmid deoxyribonucleic acid molecules in a strain of Streptoccus faecalis: Identification of a plasmid determining erythromycin resistance. J. Bacterial. 117, 283-289.

ENGEL, H. W. B., SOEDIRMAN,N., ROST, J. A., VAN LEEUWEN,W. J., AND VAN EMBDEN,J. D. A. (1980). Transferability of macrolide, lincosamide and streptogramin B resistancebetween group A, B, and D Streptococci, S. pneumoniae, and Staphylococcus aureus. J. Bacterial. 142, 407-4 13.

EVANS,R. P., AND MACRINA,F. L. (1983). Streptococcal R plasmid pIPSO1: Endonuclease site map, resistance determinant location, and construction of novel derivatives. J. Bacterial. 154, 1347- 1355. EVANS, R. P., WINTER, R. B., AND MACRINA, F. L. (1985). Molecular cloning of a pIPSO derivative yields a model replicon for the study of streptococcal conjugation. J. Gen. Microbial. 131, 145- 153. GIBSON,E. M., CHACE,N. M., LONDON,S. B., AND LQNDON, J. (1979). Transfer of plasmid-mediated antibiotic resistancefrom streptococci to lactobacilli. J. Bacteriol. 137,6 14-6 19. GONZALEZ,C. F., AND KUNKA, B. S. ( 1983). Plasmid transfer in Pediococcus spp: Intergeneric and intrageneric transfer of pIP50 1. Appl. Environ. Microbial. 46, 81-89.

HORAUD, T., BOUGUENEC,C. L., AND PEPPER,K.

SHORT

COMMUNICATIONS

( 1985). Molecular genetics of resistanceto macrolides, lincosamides streptogramin B (MLS) in streptococci. J. Antimicrob. Chemother., Suppl A 16, 11l-135. HORODNICEANU,T., BOUANCHAUD,D. H., BIETH, Cl., AND CHABBERT,Y. A. (1976). R plasmids in Streptococcusagalactiae (Group B). Antimicrob. Agents Chemother. 10,795-80 1. KRAH, E. R., III, AND MACRINA, F. L. (1989). Genetic analysis of conjugal transfer determinants encoded by the streptococcal broad host range plasmid pIPSO1. J. Bacterial. 171, 6005-6012. LEBLANC, D. L., AND LEE, L. N. (1984). Physical and genetic analysis of streptococcal plasmid pAM@l and cloning ofits replication region. J. Bacterial. 157,445453.

OULTRAM, J. D., AND YOUNG, M. (1985). Conjugal transfer of plasmid pAMP1 from Streptococcus lactis and Bacillus subtilus to Clostridium acetobutylicum. FEMS Microlett. 27, 129- 134.

69

ROMERO,D. A., SLOS,P., ROBERT,C., CASTELLINO,I., AND MERCENIER,A. (1987). Conjugative mobilization as an alternate vector delivery system for lactic streptococci. Appl. Environ. Microbial. 29,807-g 13. SCHABERG,D. R., CLEWELL,D. B., AND GLATZER,L. (1982). Conjugative transfer of R-plasmids from Streptococcusfaecalis to Staphylococcus aureus. Antimicrob. Agents. Chemother. 22, 204-207. SCHMIDHAUSER, T. J., AND HELINSKI,D. R. (1985). Regions of broad-host-range plasmid RR2 involved in replication and stable maintenance in nine speciesof gram-negative bacteria. J. Bacterial. 164,446-455. YOUNG, C., BURLAGE, R. S., AND FIGURSKI, D. H. (1987). Control of the ki/A gene of the broad-hostrange plasmid RK2: Involvement of korA, korB, and a new gene korE. J. Bacterial. 169, 13 15- 1320. Communicated by David H. Figurski